Global Self-Esteem and Specific Self-Views asfaculty.washington.edu/jdb/articles/Global Self-Esteem.pdf · Global Self-Esteem and Specific Self-Views as Determinants of People’s

Global Self-Esteem and Specific Self-Views as Determinants of People’s Reactions to Success and Failure

Keith A. Dutton and Jonathon D. Brown

University of Washington

Running head: Global Self-Esteem

May 8, 1996 Address Correspondence to: Jonathon D. Brown Department of Psychology University of Washington Box 351525 Seattle, WA 98195-1525 Phone: (206) 543-0679 Fax: (206) 685-3157 E Mail: [email protected]

Global Self-Esteem 2

Abstract A critical question in self-esteem research is whether people’s reactions to success and

failure are guided by their global self-esteem level or by their more specific beliefs about their abilities and attributes. To address this issue, the authors led participants to experience success or failure on an alleged test and then assessed their cognitive and emotional reactions to these outcomes. In Experiment 1, specific self-views predicted participants’ cognitive reactions to their performance outcomes, whereas global self-esteem predicted participants’ emotional reactions to their performance outcomes. In Experiment 2, global self-esteem predicted participants’ emotional reactions to their performance outcomes even after participants’ beliefs about their more specific abilities and attributes were taken into account. These findings suggest that when it comes to understanding people’s emotional reactions to success and failure, the effects of global self-esteem are not reducible to the way people think about their constituent qualities.


Global Self-Esteem and Specific Self-Views as Determinants of People’s Reactions to Success and Failure

Few statements are as incontrovertible as this: The way people feel about themselves is an important aspect of psychological life. Lay people and professionals alike refer to these feelings as a person’s level of self-esteem. Through the years, self-esteem has been linked to a wide variety of social psychological phenomena, including conformity (Brockner, 1984), persuasion (Rhodes & Wood, 1992), cognitive dissonance (Steele, Spencer, & Lynch, 1993), subjective well-being (Diener & Diener, 1995), and social comparison processes (Aspinwall & Taylor, 1993; Gibbons & Gerard, 1991; Wood, Giordano-Beech, Taylor, Michela, & Gaus, 1994), just to name a few.

Self-esteem also appears to influence people’s reactions to valenced outcomes. High self-esteem (HSE) people make more self-serving attributions for performance outcomes (Blaine & Crocker, 1993) and suffer less emotional distress when they fail than do low self-esteem (LSE) people (Brown & Dutton, 1995). The purpose of this article is to examine the source of these differences.

One possibility is that these tendencies stem from differences in how the two self-esteem groups think about their particular abilities and attributes. HSE people generally think they have high ability and expect to do well when undertaking an achievement-related task. These perceptions could guide their reactions to failure. After all, if one thinks one has high ability, it is only logical to assume that success is due to one’s ability but failure is not (Blaine & Crocker, 1993; Marsh, 1986; Miller & Ross, 1975). This position was articulated by Blaine and Crocker (1993).

We suggest that differences in the use of self-serving biases among high and low self-esteem individuals can be understood in terms of their differing self-concepts. In this regard, our framework is similar to cognitive interpretations of self-serving biases that have been offered by others. (Blaine & Crocker, 1993, p. 75).

Self-serving attributions, in turn, could explain why HSE people suffer less distress when they fail than do LSE people. HSE people do not assume that failure implies low ability, so they do not feel bad about themselves when they fail.

This account is certainly plausible, but there is little in the way of direct evidence to support it. To be sure, there is ample evidence that (a) HSE people think of themselves in more positive terms than do LSE people and (b) that HSE people and LSE people differ in their cognitive and emotional reactions to failure, but there are little data to connect these two sets of findings.

In fact, we are not aware of any research that has explicitly tested whether the effects of global self-esteem are due to the way people think about their more specific abilities and attributes. The closest research we could find were investigations that have examined whether self-esteem is more fruitfully viewed as a general, global construct (i.e., global self-esteem) or as a more specific belief about one’s competencies or abilities in some domain. Many prominent theorists (e.g., Bandura, 1986; Gergen, 1971; Marsh, 1990; Swann, 1990) have argued that specific self-views are the better predictor of behavior (broadly defined) than is global self-esteem, but the evidence on this point is actually mixed. Some investigations have found specific self-views to be better predictors of behavior than global self-esteem (e.g., Feather,


1969; Marsh, 1990; Swann, Pelham, & Krull, 1989), whereas other studies have found the opposite to be true or have reported mixed findings (e.g., Brockner & Hulton; 1978; Moreland & Sweeney, 1984; Shrauger & Sorman, 1977).

Shrauger’s (1975) distinction between cognitive and affective reactions to valenced feedback may shed some light on this inconsistency. Shrauger argued that expectancies of success (a specific belief about one’s ability in a particular domain) guide people’s cognitive reactions to valenced feedback, such that people with favorable expectancies regard success as more believable (and failure as less believable) than do people with unfavorable expectancies. Several investigators have tested these ideas and found support for them (e.g., Marsh, 1986; Swann, Griffin, Predmore, & Gaines, 1987). A reasonable extrapolation from this line of research is that HSE people show a greater self-serving bias than do LSE people because they think they are more able and hold higher expectancies of success.

Shrauger (1975) also examined whether expectancies of success influence people’s affective reactions to success and failure. He found that they did not: Independent of their expectations for success, people preferred success over failure. Note, however, that Shrauger used the term “affective reactions” to refer to people’s preference for positive outcomes over negative outcomes, not to people’s emotional reactions to receiving positive or negative feedback. His contention that everyone prefers success over failure does not mean that everyone experiences the same emotional reaction to success and failure.

Some of our previous research is relevant to this issue. As noted earlier, we have found that HSE people feel less humiliated and ashamed of themselves when they fail than do LSE people (Brown & Dutton, 1995). We speculated that global self-esteem, not specific self-views guided these reactions to failure, but we provided no direct evidence that this was the case. The present research allowed us to test this idea directly.

Summary and Overview of the Present Research

To summarize, there is reason to believe that specific self-views influence people’s cognitive responses to success and failure and that global self-esteem influences people’s emotional responses to success and failure. To address this issue, we had participants take a test that (allegedly) measured an important intellectual ability. Before administering the test, we assessed participants’ expectations of success and perceived ability level; afterward, we assessed their cognitive and emotional reactions to their test performance. We anticipated that specific self-views would predict what our participants thought about their test performance, but that global self-esteem would predict how they felt about their test performance.

Study 1

Method

Participants

The participants were 191 University of Washington (UW) undergraduates (63 males, 128 females).1 They participated in exchange for extra credit in various lower division psychology courses. Six additional participants did not adequately follow instructions and their data were discarded.

Materials and Procedure

The experiment was conducted in small groups of 2-4, with participants seated at separate


computers in such a way that they could not see each other’s computer screen. All instructions and experimental measures were presented on the computer.

At the start of the experiment, participants completed the Rosenberg (1965) self-esteem scale. This 10-item scale is a popular and well-validated measure of global self-esteem (Baumeister, Tice, & Hutton, 1989; Rosenberg, 1979).

After completing this measure, participants were informed that they were about to take a test of their integrative orientation ability. Integrative orientation was described as an important problem-solving ability involving the capacity to find creative and unusual solutions to problems. It was emphasized that integrative orientation was an aspect of general intelligence and was related to scholastic performance.

The experimental task was then introduced. This task was the Remote Associates Test (RAT; Mednick, 1962). With this task, participants are shown 3 words (e.g., car—swimming—cue) and asked to find a fourth word that relates to the other 3 (pool). Working interactively with the computer, participants completed 3 sample problems to ensure that they understood how the problems were solved.

The participants then learned that the test they would be taking was made up of 10 problems and that they would be given 5 minutes to solve these problems. Prior to taking the test, participants indicated (a) How well they expected to perform on the test relative to other UW students (1 = very poor; 9 = very good); (b) how their integrative orientation ability compared to other UW students (1 = bottom 10%; 9 = top 10%); and (c) how many of the 10 problems they expected to solve. Participants’ answers to these questions were highly correlated and we averaged them to form a task-specific expectancy index (alpha = .86).

The test was then administered. We varied the difficulty of the problems, so that half the participants received a set of easy problems and half received a set of difficult problems. Difficulty level was determined on the basis of prior testing with an independent sample and on published norms (McFarlin & Blascovich, 1984). Assignment to conditions was randomly determined.

After the allotted time for working on the test had expired, the computer paused for a moment and informed participants how many problems they had correctly solved. After receiving this information, participants completed two additional questionnaires (in counterbalanced order). One questionnaire assessed participants’ cognitive reactions to their test performance. This questionnaire asked participants to indicate (a) how accurately they thought the test assessed integrative orientation ability and (b) the extent to which they thought their test performance was due to their integrative orientation ability (1 = not at all; 9 = very much). Scores on the two measures were highly correlated (r = .65) and we averaged them to derive a measure of how much participants thought their test score accurately represented their ability level.

The other questionnaire assessed participants’ emotional reactions to their test performance. Here, participants indicated the extent to which they were presently experiencing each of 4 emotions (proud, pleased with myself, humiliated, ashamed) (1 = not at all; 7 = very much). These emotions were selected for study because they directly bear on how people feel about themselves after they have succeeded or failed (Brown & Dutton, 1995). After reversing the scoring for the negative emotions, we averaged the 4 items to derive a single emotion scale


(alpha = .83).

When they had finished completing these items, participants informed the experimenter that they were finished with the experiment. They were then debriefed, thanked, and excused.

Results

Preliminary Analyses

We examined the data with a series of hierarchical regression analyses.2

Self-esteem effects. Our first set of analyses used self-esteem and task performance (dummy coded: 1 = success; 2 = failure) as predictor variables. After entering these variables into the predictive equation, we added a term representing the interaction of these variables (formed by multiplying the two predictors to form a cross-product term).

Table 1 presents the results of these analyses. The first column shows that self-esteem influenced participant’s expectancy ratings: HSE participants expected to perform better than did LSE participants. These differential expectancies did not translate into better performance, however, as self-esteem did not predict the number of problems participants solved (see second full column in Table 1).

------------------------------- Insert Table 1 about here -------------------------------

The third full column in Table 1 shows the results for participant’s cognitive reactions to their task performance. The significant cross-product term (b = -.15) means that self-esteem and task performance interacted to influence these judgments. To examine the nature of the interaction, we calculated predicted values for participants scoring one standard deviation above and below the mean on each variable (Cohen & Cohen, 1983). As shown in Figure 1, LSE participants regarded success and failure as equally diagnostic of their ability level, but HSE participants thought the test provided a more accurate assessment of their ability when they did well than when they did poorly. Another way of saying this is that HSE participants exhibited a self-serving bias, but LSE participants did not.

------------------------------- Insert Figure 1 about here -------------------------------

The final full column in Table 1 shows the results for the analysis of emotion. The significant cross-product term (b = .19) again means that self-esteem and task performance interacted to predict these reactions. Figure 2 reveals the nature of the interaction. Consistent with prior research (Brown & Dutton, 1995), the figure shows that HSE participants experienced less emotional distress following failure than did LSE participants.

------------------------------- Insert Figure 2 about here -------------------------------

Expectancies. In our next set of analyses, we substituted pretask expectancies for self-esteem (see Table 2). Although expectancies did not predict participant’s emotional reactions to their test performance, expectancies did interact with performance outcomes to predict participants’ ability attributions (b = -.19). Figure 3 shows the nature of the interaction. As can be seen, participants with high expectancies of success made more self-serving attributions for their test performance than did participants with low expectancies of success. The form of the


interaction is virtually identical to the one found when self-esteem was used as a predictor. ------------------------------- Insert Table 2 about here ------------------------------- ------------------------------- Insert Figure 3 about here -------------------------------

Main Analyses

The preceding results set the stage for our main analyses. To this point, we have seen that global self-esteem and specific self-views (i.e., pretask expectancies) each interact with performance outcomes to predict participant’s cognitive reactions to success and failure, but that only global self-esteem interacts with performance outcomes to predict participant’s emotional reactions to success and failure. The key issue now is whether the effects of global self-esteem are due to participant’s more specific beliefs about their abilities.

To explore this issue, we conducted a hierarchical regression analysis that included self-esteem, pretask expectancies, and performance outcomes as predictor variables (as well as all possible two-way and three-way interactions, entered in a step-wise fashion). The logic behind this approach is as follows: If differential expectancies explain why HSE participants made more self-serving attributions than did LSE participants, we should find that this effect disappears once expectancies have been statistically removed. A regression analysis with both predictors in the equation allows us to test this hypothesis, because the analysis simultaneously tests the independent effect of all predictor variables (see Cohen & Cohen, 1983 for a more complete discussion of this issue).

Inspection of the first full column 1 in Table 3 shows the results with ability attributions as the criterion variable. The lack of a significant self-esteem X outcome interaction (b = -.09) means that once expectancies were statistically controlled, HSE participants were no more self-serving in their causal judgments than were LSE participants. Note, however, that the reverse was not true. Even after the effects of self-esteem were statistically controlled, expectancies continued to interact with performance outcomes to predict attributions (b = -.16). Taken together, these findings imply that self-esteem differences in cognitive reactions to performance feedback are cognitively-mediated. HSE people think they have high ability and expect to succeed, and this explains why they make more self-serving attributions than do LSE people (Blaine & Crocker, 1993; Brockner et al. 1993; Marsh, 1986).


But what of participant’s emotional reactions to their performance outcomes? Can differential expectancies explain why HSE people are less adversely affected by failure than are LSE people? To answer this question, we repeated our analyses using emotional reactions as the dependent variable. The second full column in Table 3 shows that even after the effects of expectancies were statistically removed, self-esteem continued to interact with task performance to predict participant’s emotional responses (b = .20). As before, the effect was such that HSE participants were less adversely affected by failure than were LSE participants. This finding suggests that self-esteem differences in emotional reactions to performance outcomes are not


driven by people’s more specific beliefs about their ability.

Supplementary Analyses

Cognitive theories of emotion assert that people’s emotional responses to success and failure are governed by their beliefs about the causes of their performance (e.g., Weiner, 1986). This view suggests that attributions may mediate self-esteem differences in emotion. Expectancies determine attributions, and attributions determine emotion.

To address this issue, we performed another regression analysis, substituting the attributions participants made for their performance outcomes for pretask expectancies. The results of this analysis are shown in Table 4. The data reveal two interesting effects. First, attributions and outcome interacted (b = -.16) to predict emotion in the manner specified by Weiner (1986). This interaction did not, however, alter the strength of the self-esteem X outcome interaction, which remained significant (b = .22). The fact that self-esteem continued to interact with task performance to predict emotion, even after the effects of attributions were statistically removed, means that differences in ability attributions do not explain why HSE participants were less adversely affected by failure than were LSE participants.


Discussion

Study 1 examined whether self-esteem differences in response to failure are due to the way people think about their more specific attributes. Evidence supporting this hypothesis was found when we examined participant’s cognitive reactions to success and failure. The tendency for HSE participants to make more self-serving attributions for their task performance than LSE participants was eliminated once we statistically controlled for participant’s pretask expectancies of success and beliefs about their ability. This finding supports the claim that the reason why HSE people make more self-serving judgments is because they believe they are more able (Blaine & Crocker, 1993; Brockner et al., 1993; Marsh, 1986).

There was no evidence, however, that specific self-views mediate self-esteem differences in emotional reactions. Regardless of their pretask expectancies, and independent of the attributions they made for their performance, HSE people felt less distressed when they failed than did LSE participants. This finding suggests that when it comes to people’s emotional reactions to failure, the effects of global self-esteem are not reducible to the way people think about their more constituent qualities.

At this point, this conclusion should be regarded as preliminary and in need of further clarification. For one thing, pretask expectancies may not be the most relevant cognitive variable. People’s beliefs about their attributes and abilities across a variety of domains might be a more important determinant of their emotional reactions to success and failure.

This conjecture follows from Steele’s (1988) self-affirmation theory. Steele has argued that people cope with negative outcomes in one domain by focusing on their virtues in other, unrelated domains (see also, Baumeister, 1992; Brown & Dutton, 1995; Brown & Smart, 1991; Greenberg & Pyszczynski, 1985). When applying these ideas to understanding self-esteem effects, Steele has argued that HSE people are better insulated from the negative effects of failure because they believe they have so many other fine qualities (Josephs, Larrick, Steele, & Nisbett,


1992; Steele et al., 1993). Presumably, these perceived qualities function as cognitive resources, which HSE people call upon when they fail.

From this perspective, specific self-views (i.e., people’s beliefs about their specific competencies, attributes, and abilities) explain self-esteem differences in emotional reactions to success and failure. HSE people believe they have many more positive qualities than do LSE people, and this is the reason why they experience less distress when they fail. Although there is some evidence to support this claim (Steele et al., 1993), it is largely indirect. To our knowledge, no investigation has (a) measured people’s beliefs about their attributes in a variety of areas, (b) exposed people to failure, and (c) then determined whether attribute-specific beliefs mediate self-esteem differences in response to a negative outcome. Study 2 was designed to provide such a test.

Study 2

Method

Participants

The participants were 136 UW undergraduates (42 males, 94 females) who participated in exchange for extra credit in various lower division psychology courses. The data from four additional participants were discarded. Two expressed suspicion regarding their test performance and two did not follow instructions.

Materials and Procedure

There was one important difference between this study and Study 1. Instead of assessing participant’s perceived competence at the experimental task, we examined their more general perceived qualities. This was accomplished by having the participants complete a trait questionnaire at the start of the experimental session, just after filling out the Rosenberg (1965) self-esteem scale. The questionnaire asked them to indicate how well 10 attributes described them (1 = not at all; 5 = very much). Five of the attributes were positive (intelligent, athletic, attractive, talented, and kind) and five were negative (unintelligent, uncoordinated, unattractive, incompetent, and inconsiderate). These items were chosen because they are highly important to college students (see, for example, Pelham & Swann, 1989).

Results and Discussion

Preliminary Analyses

We first reversed the scoring of the 5 negative self-evaluation items and then summed the 10 items to form a self-evaluation index (alpha = .72). Scores on this index were highly related to self-esteem (r = 62, p < .001), supporting the claim that HSE people evaluate themselves more positively than do LSE people.

Next, we conducted a regression analysis to determine whether self-esteem interacted with the success/failure manipulation in the prediction of emotion. The predictors were (standardized) self-esteem scores and a dummy variable representing the success/failure manipulation (1 = success; 2 = failure). An interaction term was then created by calculating the cross-product of the two (standardized) variables. The top-half of Table 5 shows the results. As can be seen, all three sources of variance were significant (all ps < .01). Inspection of the interaction by graphing (see left-hand side of Figure 4) indicated that HSE participants again reported less emotional distress following failure than did LSE participants.


------------------------------- Insert Table 5 about here ------------------------------- ------------------------------- Insert Figure 4 about here -------------------------------

Next, we repeated our analyses substituting scores on the self-evaluation index for self-esteem (see bottom-half of Table 5). All three sources of variance (self-evaluation, outcome, and the interaction of these two variables) were significant (all ps < .05), indicating that scores on the self-evaluation measure also moderated the effects of performance outcomes on emotion. The right-hand side of Figure 4 shows that the effect closely paralleled the one observed for self-esteem.

Main Analyses

Thus far, the data show that both global self-esteem and the manner in which people evaluate their specific characteristics predict their emotional responses to success and failure. The critical issue now is whether the effects of global self-esteem depend on how people evaluate their specific characteristics. In particular, are HSE people less adversely affected by failure because they think they have many more positive qualities than do LSE people?

To answer this question, we conducted a regression analysis that included all three predictors (self-esteem, self-evaluations, and outcome) in a single equation (as well as all two-way interactions and the three-way interaction, entered in a step-wise fashion). Table 6 shows the results. It can be seen that the self-esteem X outcome interaction remained significant (b = .18), but the self-evaluation X outcome interaction did not (b = .07). In short, global self-esteem moderated participants’ emotional reactions to their performance outcomes, and this effect did not depend on whether or not participants thought they were good or bad at things (i.e., intelligent, talented, unattractive, or uncoordinated).


Supplemental Analyses

We conducted a number of ancillary analyses to determine whether the preceding results were reliable across the various attributes we measured. First, we formed scales by averaging 2 traits in each of 5 domains: intelligence (intelligent, unintelligent); social qualities (kind, inconsiderate); attractiveness (attractive, unattractive); athleticism (athletic, uncoordinated), and competence (talented, incompetent). We then standardized these scores and substituted each of them (one at a time) for the general self-evaluation index used in the main analyses. In all 5 cases, the self-esteem X outcome interaction remained a significant predictor of emotion after controlling for these more specific self-evaluations.

In another analysis, we first calculated whether participant’s ratings fell above or below the mean in each of these five domains. We then tallied the number of domains in which the participants rated themselves more positively than their peers. Scores on this index could range from 0 (indicating that a given participant rated himself or herself below the mean in all 5 areas) to 5 (indicating that a given participant rated himself or herself above the mean in all 5 areas). Insofar as these scores reflect the number of areas in which participants believe they excel, they may provide the fairest test of Steele et al.’s (1993) cognitive resources model.


Scores on this index were correlated with self-esteem (r = .55, p < .001), and they did interact with performance outcomes to predict participant’s emotional responses to success and failure when self-esteem was not in the equation (b = .17, p < .05). However, when self-esteem was included in the equation, the only interaction to predict emotion was the self-esteem X outcome interaction (b = .19, p < .05).

In sum, our effects generalize across the various attributes we assessed. HSE people feel less distress when they fail than do LSE people and this is not simply because they think they have more positive traits.

General Discussion

The term self-esteem has been used in multiple ways. Global self-esteem refers to the way people generally feel about themselves; attribute specific self-esteem refers to the way people evaluate their specific attributes and abilities (e.g., people who believe they are good in math are said to have HSE in that domain of life). In this article, we explored the role these constructs play in guiding people’s cognitive and emotional reactions to success and failure. We found that specific self-views predict people’s cognitive reactions to performance outcomes, but global self-esteem predicts people’s emotional reactions to performance outcomes.

Potential Limitations

Before discussing the implications of these findings, we feel obliged to alert readers to some potential limitations. The analytic strategy we used, in which we statistically controlled for one variable while examining the effects of another does not provide a strong form of inference. Various factors, including measurement error and differences in range can explain why the effects of one variable are stronger than another. Clearly, it would have been preferable to experimentally manipulate the variables of interest here, although it is difficult to see how that could be done for global self-esteem and people’s general beliefs about their abilities and attributes.

At the same time, it is worth noting that expectancies did mediate self-esteem differences in Study 1. The tendency for HSE people to make more self-serving attributions for their performance was eliminated once expectancies were taken into account. The fact that the expectancy measure was good enough to establish this effect lessens the likelihood that the effects we observed elsewhere are due to methodological limitations.

Another possibility is that we didn’t measure the right attributes in Study 2. Perhaps our measure wasn’t broad enough. Maybe we should have assessed participant’s beliefs about their musical ability, cooking skills, or any of a dozen other attributes and abilities. We could also have included a measure of how important these self-views were or how certain participants were of their standing on each attribute (Campbell, 1990; Pelham, 1991). These are important issues that our research failed to capture.

Theoretical and Practical Implications

These limitations notwithstanding, we believe our results reveal some important details about the nature of self-esteem. HSE people do think they have many more positive qualities than do LSE people, but these differences do not appear to tell the whole story. Our data indicate that even LSE people who think they have many positive qualities feel humiliated and ashamed of themselves when they fail. Why might this be the case?


We believe the most likely possibility is that feelings of humiliation and shame are not responses to judgments about specific attributes and qualities. They reflect a more general and undifferentiated feeling that one is a bad person, rather than simply being bad at things (Sullivan, 1953). In our opinion, these feelings comprise the essence of LSE. Indeed, we believe it is entirely possible for a LSE person to say “Yes, I know I am smart and can do many things well, but I just don’t feel good about myself.” These perceptions surface when failure is encountered, evoking feelings of shame and humiliation.

Once activated, these feelings may also spread and color the way LSE people think about their specific attributes (Brown & Mankowski, 1993). Several investigations have found that LSE people overgeneralize the negative implications of failure. Failure in one domain leads them to evaluate themselves more negatively in other domains (Brown & Dutton, 1995; Brown & Smart, 1991; Kernis, Brockner, & Frankel, 1989). HSE people tend to do just the opposite. They tend to compensate for failure by exaggerating their virtues in alternative domains (Brown & Smart, 1991).

At first blush, these opposing tendencies seem inconsistent with our claim that people’s emotional reactions to failure depend on their global self-esteem level, not on how they evaluate their specific qualities. In this regard, it is important to note that overgeneralization effects obtain after failure has already occurred. Consequently, they represent different ways of coping with failure, rather than static differences in how people typically evaluate their specific attributes.

An examination of the way LSE people evaluate their specific characteristics supports this characterization. Within normal populations, LSE people do not regard themselves in excessively negative terms (Baumeister et al., 1989). They do not believe they are stupid, homely, or clumsy. Instead, they generally hold positive beliefs about themselves, particularly relative to their beliefs about others (Brown, 1986, 1993; Brown & Gallagher, 1992). Not as positive as HSE people, of course, but positive nonetheless. This pattern emerged in the present research as well. Even participants who scored in the bottom 10% of the self-esteem distribution in Study 2 rated themselves above the scale midpoint of 3 across the 10 attributes (M = 3.40). This value was significantly greater than the scale midpoint, t(16) = 2.67, p < .05.

The absence of true negativity on the part of LSE people suggests that negative beliefs alone do not explain self-esteem differences in behavior. It doesn’t make sense to claim that LSE are so adversely affected by failure because they think they can’t do anything well and have many negative qualities, when in fact they regard themselves as reasonably intelligent and talented, and not at all unattractive and incompetent.

It might be argued that college student samples account for the lack of clear negativity LSE people show. We are not aware of any evidence that college students score higher in self-esteem than other groups, but we do not doubt that true self-deprecation occurs in some populations (e.g., the severely depressed). This admission does not negate the point we are making here, which is this: self-esteem differences are reliably found in the absence of absolute negativity on the part of participants classified as having LSE. It must be because the observed effects are not due to negative thinking per se.

Our findings also bear on how self-esteem should be measured. Some scales assess self-esteem by aggregating people’s views of themselves across different domains (e.g., Coopersmith, 1967). The fact that specific self-views and global self-esteem have different


correlates suggests that this strategy should be avoided. We are equally skeptical of research that claims to manipulate self-esteem by giving people false feedback regarding some aspect of themselves (e.g., Heatherton & Polivy, 1991). Our findings cast doubt on the assumption that global self-esteem can be mimicked by leading people to believe they are high or low in some ability, or are good or bad in some domain.

Our findings also speak to interventions designed to instill self-esteem. Many of these programs try to raise self-esteem by encouraging people to focus on their positive qualities, to believe they can do many things well. Approaches like these are predicated on the dual assumptions (a) that self-esteem is built from the bottom-up (from specific beliefs to global self-esteem) and (b) that negative self-relevant beliefs are the defining feature of LSE. We see good reason to question each of these assumptions. As noted earlier, many LSE people already believe they have many fine qualities, but these beliefs do not protect them from the pain of failure. These findings suggest that specific self-beliefs, however positive they may be, are not the key to having HSE.

This naturally leads one to wonder just what is the key to having HSE. Unfortunately, the present research is better suited to showing what HSE isn’t than establishing what it is. Elsewhere (Brown, 1993; Brown & Dutton, 1995), we have argued that the essence of HSE is an unconditional feeling of affection for oneself that does not depend on the perception that one has any particular positive quality or qualities (Rogers, 1951). We have likened this feeling to the way most (though not all) parents feel toward their children. Most parents feel an abiding love for their children that is not contingent on whether their children are “good at things.” Somehow, HSE people introject these feelings early on in life and thereafter do not judge themselves by what they can or cannot do. Epstein (1980) reached a similar conclusion:

People with HSE, in effect, carry with them a loving parent who is proud of their successes and tolerant of their failures. ... Although capable of being disappointed and depressed by specific experiences, people with HSE recover quickly, as do children who are secure in their mother’s love. In contrast, people with LSE carry within them a disapproving parent who is harshly critical of their failures, and register only short-lived pleasures when they succeed. Such people are apt to be unduly sensitive to failure and to rejection, to have low tolerance for frustration, to take a long time to recover following disappointments, and to have a pessimistic view of life. The picture is not unlike that of children who are insecure in their parent’s love. (p. 106)

Concluding Remarks

Broad personality variables once dominated psychological research (Allport, 1937). But this is no longer the case. Beginning with Mischel’s (1968) critique, many psychologists turned away from global aspects of personality toward more specific, cognitively-based constructs (Cantor & Kihlstrom, 1981). In a similar vein, social psychologists working in the attitude area have argued that specific beliefs and attitudes are better predictors of behavior than are more general beliefs and attitudes (Ajzen & Fishbein, 1977).

This emphasis on specificity over generality also found its way into research on self-concept. Many contemporary theorists focus their research on specific views of the self rather than on global self-esteem (e.g., Bandura, 1986; Markus & Wurf, 1987; Swann, 1990). Some have even gone so far as to claim that global self-esteem is a fiction (Gergen, 1971) or of limited


predictive value (Marsh, 1990).

The present findings suggest an important qualification to these conclusions. Although specific self-views may often be a better predictor of behavior than is global self-esteem, this is not the case when it comes to understanding how people feel about themselves when they fail. Insofar as many important tasks in life involve overcoming obstacles and dealing with set backs, rejection, and disappointment, global self-esteem would seem to play a critical role in psychological life.


References Ajzen, I., & Fishbein, M. (1977). Attitude-behavior relations: A theoretical analysis and review

of empirical research. Psychological Bulletin, 84, 888-918.

Allport, G. W. (1937). Personality: A psychological interpretation. New York: Holt, Rinehart, & Winston.

Aspinwall, L. G., & Taylor, S. E. (1993). Effects of social comparison direction, threat, and self-esteem on affect, self-evaluation, and expected success. Journal of Personality and Social Psychology, 64, 708-722.

Bandura, A. (1986). Social foundations of thought and action. Englewood Cliffs, NJ: Prentice-Hall.

Baumeister, R. F. (1982). Self-esteem, self-presentation, and future interaction: A dilemma of reputation. Journal of Personality, 50, 29-45.

Baumeister, R. F., Tice, D. M., & Hutton, D. G. (1989). Self-presentational motivations and personality differences in self-esteem. Journal of Personality, 57, 547-579.

Blaine, B., & Crocker, J. (1993). Self-esteem and self-serving biases in reactions to positive and negative events: An integrative review. In R. F. Baumeister (Ed.), Self-esteem: The puzzle of low self-regard (pp. 55-85). New York: Plenum Press.

Brockner, J. (1984). Low self-esteem and behavioral plasticity: Some implications for personality and social psychology. In L. Wheeler (Ed.), Review of personality and social psychology (Vol. 4, pp. 237-271). Beverly Hills: Sage.

Brockner, J., & Hulton, A. J. B. (1978). How to reverse the vicious cycle of low self-esteem: The importance of attentional focus. Journal of Experimental Social Psychology, 14, 564-578.

Brockner, J., Wiesenfeld, B. M., & Raskas, D. F. (1993). Self-esteem and expectancy-value discrepancy: The effects of believing that you can (or can’t) get what you want. In R. F. Baumeister (Ed.), Self-esteem: The puzzle of low self-regard (pp. 219-240). New York: Plenum Press.

Brown, J. D. (1986). Evaluations of self and others: Self-enhancement biases in social judgments. Social Cognition, 4, 353-376.

Brown, J. D. (1993). Self-esteem and self-evaluation: Feeling is believing. In J. Suls (Ed.), Psychological perspectives on the self (Vol. 4, pp. 27-58). Hillsdale, NJ: Lawrence Erlbaum Associates.

Brown, J. D., & Dutton, K. A. (1995). The thrill of victory, the complexity of defeat: Self-esteem and people’s emotional reactions to success and failure. Journal of Personality and Social Psychology, 68, 712-722.

Brown, J. D., & Gallagher, F. M. (1992). Coming to terms with failure: Private self-enhancement and public self-effacement. Journal of Experimental Social Psychology, 28, 3-22.


Brown, J. D., & Mankowski, T. A. (1993). Self-esteem, mood, and self-evaluation: Changes in mood and the way you see you. Journal of Personality and Social Psychology, 64, 421-430.

Brown, J. D., & Smart, S. A. (1991). The self and social conduct: Linking self-representations to prosocial behavior. Journal of Personality and Social Psychology, 60, 368-375.

Campbell, J. D. (1990). Self-esteem and clarity of the self-concept. Journal of Personality and Social Psychology, 59, 538-549.

Cantor, N. & Kihlstrom, J. F. (Eds.). (1981). Personality, cognition, and social interaction. Hillsdale, NJ: Lawrence Erlbaum Associates.

Cohen, J., & Cohen, P. (1983). Applied multiple regression/correlation analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Lawrence Erlbaum Associates.

Coopersmith, S. (1967). The antecedents of self-esteem. San Francisco: W. H. Freeman.

Diener, E., & Diener, M. (1995). Cross-cultural correlates of life satisfaction and self-esteem. Journal of Personality & Social Psychology, 68, 653-663.

Epstein, S. (1980). The self-concept: A review and the proposal of an integrated theory of personality. In E. Staub (Ed.), Personality: Basic issues and current research (pp. 82-132) Englewood Cliffs, NJ: Prentice-Hall.

Feather, N. T. (1969). Attribution of responsibility and valence of success and failure in relation to initial confidence and task performance. Journal of Personality and Social Psychology, 13, 129-144.

Gergen, K. J. (1971). The concept of self. New York: Holt, Rinehart, & Winston.

Gibbons, F. X., & Gerrard, M. (1991). Downward comparison and coping with threat. In J. Suls & T. A. Wills (Eds.), Social comparison: Contemporary theory and research (pp. 317-345). Hillsdale, NJ: Lawrence Erlbaum Associates.

Greenberg, J., & Pyszczynski, T. (1985). Compensatory self-inflation: A response to the threat to self-regard of public failure. Journal of Personality and Social Psychology, 49, 273-280.

Heatherton, T. F., & Polivy, J. (1991). Development and validation of a scale for measuring state self-esteem. Journal of Personality and Social Psychology, 60, 895-910.

Josephs, R. A., Larrick, R. P., Steele, C. M., & Nisbett, R. E. (1992). Protecting the self from the negative consequences of risky decisions. Journal of Personality and Social Psychology, 62, 26-37.

Kernis, M. H., Brockner, J., & Frankel, B. S. (1989). Self-esteem and reactions to failure: The mediating role of overgeneralization. Journal of Personality and Social Psychology, 57, 707-714.

Markus, M. & Wurf, E. (1987). The dynamic self-concept: A social psychological perspective. Annual Review of Psychology, 38, 299-337.

Marsh, H. W. (1986). The self-serving effect (bias?) in academic attributions: Its relation to academic achievement and self-concept. Journal of Educational Psychology, 78, 190-200.


Marsh, H. W. (1990). A multidimensional, hierarchical model of self-concept: Theoretical and empirical justification. Educational Psychology Review, 2, 77-172.

McFarlin, D. B., & Blascovich, J. (1984). On the Remote Associates Test (RAT) as an alternative to illusory performance feedback: A methodological note. Basic and Applied Social Psychology, 5, 223-229.

Mednick, S. A. (1962). The associative basis of the creative process. Psychological Review, 26, 220-232.

Mischel, W. (1968). Personality and assessment. New York: Wiley.

Miller, D. T., & Ross, M. (1975). Self-serving biases in the attribution of causality: Fact or fiction? Psychological Bulletin, 82, 213-235.

Moreland, R. L., & Sweeney, P. D. (1984). Self-expectancies and reactions to evaluations of personal performance. Journal of Personality, 52, 156-176.

Pelham, B. W. (1991). On confidence and consequences: The certainty and importance of self-knowledge. Journal of Personality and Social Psychology, 60, 518-530.

Pelham, B. W., & Swann, W. B., Jr. (1989). From self-conceptions to self-worth: On the sources and structure of global self-esteem. Journal of Personality and Social Psychology, 57, 672-680.

Rhodes, N., & Wood, W. (1992). Self-esteem and intelligence affect influenceability: The mediating role of message reception. Psychological Bulletin, 111, 156-171.

Rogers, C. R. (1951). Client-centered therapy. Boston: Houghton Mifflin.

Rosenberg, M. (1965). Society and the adolescent self-image. Princeton, NJ: Princeton University Press.

Rosenberg, M. (1979). Conceiving the self. New York: Basic Books.

Shrauger, J. S. (1975). Responses to evaluation as a function of initial self-perceptions. Psychological Bulletin, 82, 581-596.

Shrauger, J. S., & Sorman, P. B. (1977). Self-evaluations, initial success and failure, and improvement as determinants of persistence. Journal of Consulting and Clinical Psychology, 45, 784-795.

Steele, C. M. (1988). The psychology of self-affirmation: Sustaining the integrity of the self. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 21, 261-302). New York: Academic Press.

Steele, C. M., Spencer, S. J., & Lynch, M. (1993). Self-image resilience and dissonance: The role of affirmational resources. Journal of Personality and Social Psychology, 64, 885-896.

Sullivan, H. S. (1953). The interpersonal theory of psychiatry. New York: W. W. Norton.

Swann, W. B., Jr. (1990) To be adored or to be known? The interplay of self-enhancement and self-verification. In R. M. Sorrentino & E. T. Higgins (Eds.), Motivation and cognition (Vol. 2, pp. 408-448). New York: Guilford Press.


Swann, W. B., Jr., Griffin, J. J., Predmore, S. C., & Gaines, B. (1987). The cognitive-affective crossfire: When self-consistency confronts self-enhancement. Journal of Personality and Social Psychology, 52, 881-889.

Swann, W. B., Jr., Pelham, B. W., & Krull, D. S. (1989). Agreeable fancy or disagreeable truth? Reconciling self-enhancement and self-verification. Journal of Personality and Social Psychology, 57, 782-791.

Wood, J. V., Giordano-Beech, M., Taylor, K. L., Michela, J. L., & Gaus, V. (1994). Strategies of social comparison among people with low self-esteem: Self-protection and self-enhancement. Journal of Personality and Social Psychology, 67, 713-731.

Weiner, B. (1986). An attributional theory of motivation and emotion. New York: Springer.


Author Identification Notes This research was supported by a Presidential Young Investigator Award from the

National Science Foundation (BNS-8958211) and by a grant from the Horizons Foundation of Seattle.


Footnotes 1Sex of subject did not modify any of the findings in this study or in Study 2, and will not

be discussed further. 2In all regression analyses reported in this paper, all variables were standardized prior to

conducting the analyses.


Table 1

Hierarchical Multiple Regression Analyses Relating Self-Esteem and Outcome to Various Criterion Measures: Study 1

Criterion _________________________________________________________________________ Pretask Task Ability Expectancies Performance Attributions Emotion ________________________________________________________________________ Predictor b ∆R2 b ∆R2 b ∆R2 b ∆R2

Self-Esteem .33*** .03 .11 .36*** Outcome -.01 .11*** -.71*** .50*** -.20** .05** -.40*** .28*** Self-Esteem X Outcome .04 .00 .04 .00 -.15* .02* .19** .04** Note: Entries are unstandardized regression coefficients. R2 values are increments in variance for each step. * p < .05; ** p < .01; *** p < .001


Table 2

Hierarchical Multiple Regression Analyses Relating Expectancies and Outcome to Various Criterion Measures: Study 1

Criterion _______________________________________ Ability Attributions Emotion _______________________________________ Predictor b ∆R2 b ∆R2 Expectancies .04 .06 Outcome -.20** .04* -.39*** .15*** Expectancies X Outcome .19** .03** .05 .00 Note: Entries are unstandardized regression coefficients. R2 values are increments in variance for each step. * p < .05; ** p < .01; *** p < .001


Table 3

Hierarchical Multiple Regression Analyses Relating Self-Esteem, Outcome, and Expectancies to Various Criterion Measures: Study 1

Criterion _______________________________________ Ability Attributions Emotion _______________________________________ Predictor b ∆R2 b ∆R2 Self-Esteem .10 .38*** Outcome -.20** -.40** Expectancies .00 .05* -.07 .28*** Self-Esteem X Outcome -.09 .20** Expectancies X Outcome -.16* -.04 Self-Esteem X Expectancies .02 .05* -.03 .04* Self-Esteem X Expectancies X Outcome -.08 .00 .06 .00 Note: Entries are unstandardized regression coefficients. R2 values are increments in variance for each step. * p < .05; ** p < .01; *** p < .001


Table 4

Hierarchical Multiple Regression Analyses Relating Self-Esteem, Outcomes, and Attributions to Emotion: Study 1

Criterion ________ Emotion ________ Predictor b ∆R2

Self-Esteem .35*** Outcome -.39*** Attributions .05 .28*** Self-Esteem X Outcome .22*** Attributions X Outcome -.16** Self-Esteem X Attributions .01 .06*** Self-Esteem X Attributions X Outcome .12 .01 Note: Entries are unstandardized regression coefficients. R2 values are increments in variance for each step. * p < .05; ** p < .01; *** p < .001


Table 5

Hierarchical Multiple Regression Analyses Relating Global Self-Esteem and Outcomes (top-half) and Specific Self-Evaluations and Outcomes (bottom-half) to Emotion: Study 2


Self-Esteem .39*** Outcome -.35*** .27*** Self-Esteem X Outcome .23** .05** Self-Evaluations .28*** Outcome -.38*** .20*** Self-Evaluations X Outcome .18* .03* Note: Entries are unstandardized regression coefficients. R2 values are increments in variance for each step. * p < .05; ** p < .01; *** p < .001


Table 6

Hierarchical Multiple Regression Analyses Relating Global Self-Esteem, Specific Self-Evaluations, and Outcomes to Emotion: Study 2


Self-Esteem .34*** Outcome -.36*** Self-Evaluations .07 .27*** Self-Esteem X Outcome .18* Self-Evaluations X Outcome .07 Self-Esteem X Self-Evaluations .00 .05* Self-Esteem X Outcome X Self-Evaluations .00 .00 Note: Entries are unstandardized regression coefficients. R2 values are increments in variance for each step. * p < .05; ** p < .01; *** p < .001



Figure Captions

Figure 1. Schematic representation of the interaction between self-esteem and performance outcomes in the prediction of ability attributions: Study 1

Figure 2. Schematic representation of the interaction between self-esteem and performance outcomes in the prediction of emotion: Study 1

Figure 3. Schematic representation of the interaction between expectancies and performance outcomes in the prediction of ability attributions: Study 1

Figure 4. Schematic representation of the interaction between self-esteem and performance outcomes (left-hand side of figure) and between self-evaluations and performance outcomes (left-hand side of figure) in the prediction of emotion: Study 2


Documents

Global Self-Esteem and Specific Self-Views asfaculty.washington.edu/jdb/articles/Global Self-Esteem.pdf · Global Self-Esteem and Specific Self-Views as Determinants of People’s